Near-identical pages made to catch searches

How to do it

Google's spam policies call this doorway abuse. Its examples are near-identical domains or home pages made for one query, and city or region pages that funnel visitors to one page. Pages that exist only to lead into the useful part of a site count too. A page with facts specific to its town is fine, as Give a place its own page only with local content says.

Examples

JavaScript example

const shingles = text => new Set(text.toLowerCase().split(/\s+/).map((word, i, words) => words.slice(i, i + 3).join(' ')));
const overlap = (a, b) => [...a].filter(shingle => b.has(shingle)).length / Math.min(a.size, b.size);
for (const [first, second] of pairsFromSameTemplate) assert.ok(overlap(shingles(first.text), shingles(second.text)) <= 0.8, `${first.path} and ${second.path} are near copies`);

Python example

def shingles(text):
    words = text.lower().split()
    return {" ".join(words[i:i + 3]) for i in range(len(words))}

for first, second in pairs_from_same_template:
    a, b = shingles(first.text), shingles(second.text)
    assert len(a & b) / min(len(a), len(b)) <= 0.8, f"{first.path} and {second.path} are near copies"

How to check

A build test compares the main text of every pair of pages made from the same template. No pair shares more than 80% of its three-word sequences.

Sources

• Google Search Central – Spam policies, doorway abuse

Related rules

• SEO-64 Give a place its own page only with local content in SEO